Music player CPU usage? - Verizon Droid Charge

So i was playing music today (first time in a while) and not doing much else. I noticed the battery was draining rapidly so here is CPU spy:
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
I suspect most of that 1200 is the music. How can i keep it down? I am using the stock music player btw
Sent from my pocket-sized, Linux-based computer using electromagnetic radiation... and Tapatalk.

I've found if you go into the market and type in android music you will get the stock froyo music player. It sounds the same and uses about 1/4 as much ram as the stock player
Sent from my SCH-I510 using XDA App

Go to you're about phone in settings and look at battery use. If music is what's causing the battery loss, you can stop overclocking and undervolt at 800 and 1000 aggressively and that should help while you're listening to music.

What CPU governor are you using?

HTCluva said:
I've found if you go into the market and type in android music you will get the stock froyo music player. It sounds the same and uses about 1/4 as much ram as the stock player
Sent from my SCH-I510 using XDA App
Click to expand...
Click to collapse
Thanks; I'll try this.
hazard209 said:
Go to you're about phone in settings and look at battery use. If music is what's causing the battery loss, you can stop overclocking and undervolt at 800 and 1000 aggressively and that should help while you're listening to music.
Click to expand...
Click to collapse
I don't know why, but UV has never worked properly for me. The phone always restarts itself when the CPU enters a UV'd state. Maybe i just need specifics on how to do it.
imnuts said:
What CPU governor are you using?
Click to expand...
Click to collapse
I am using smartassv2. Would conservative help?

Well start simple and test UV values. Just do -50 across the board to start.
Here's what I use: IO Scheduler - noop, CPU Governor - smartassV2, 1200/-50, 1000/-50, 800/-100, 400/-100, 200/-125, 100/-125
But those are pretty aggressive UV settings.

Firstly, I suggest trying the free Mortplayer or the non-free Poweramp. I like them both.
I have the stock music player frozen and I didn't test it. I did test Poweramp and Mortplayer with 160kbps MP3 as well as uncompressed WAV while virtually nothing was running on my phone.
I found the results with OS Monitor. I stared at OS Monitor for quite a while without music playback first to make sure stuff doesn't jack the CPU. Android System jumps from 0 - 1% every now and then. System jumps from 0 - 2 % every now and then. All EQ and Tone filters were disabled.
Each of the following OS Monitor values were averaged in my mind while playing music for a few minutes for each playback type.
Poweramp 160kbps MP3: System and Poweramp processes together averaged 12% (each process was roughly the same average, so the load was split)
Poweramp 44.1 kHz 16bit stereo WAV: System and Poweramp processes together averaged about 9% (Poweramp process was between 1 and 2%- System took most of the load.)
Mortplayer 160kbps MP3: Mortplayer process 0% with System taking the load at 12% average.
Mortplayer 44.1 kHz 16bit stereo WAV: Mortplayer process 0% with System taking the load at 9% average.
In conclusion, Mortplayer and Poweramp use the same overall CPU resources during playback. Playback of MP3 uses more CPU resources than uncompressed WAV files. Mortplayer uses the System process while Poweramp shares the load with the System process.
My setup.
IMO 4.0 kernel with Tegrak Overclock with Interactive X governor. Overclocked to 1.252Ghz. Minimum scaling frequency is at 200Mhz.
CPU Spy shows that I use multiple CPU states (mostly 200, my minimum scaling frequency) during playback, as evidenced by resetting the timers in CPU Spy and refreshing after some time has lapsed while playing music. Is there a way to see the active state? My guess is that the state changes in microseconds, so the existing state status would be pointless (just a guess.)
Does anybody know if OS Monitor is calculating the CPU usage with respect to the maximum frequency all the time? I suspect this is the case.
Sorry for the derailing action.

Related

SETCPU

I've download SETCPU and for some reason, I don't see the phone listed.
NVIDIA tegra 2 TP20/AP2OH?
t-mobile Pulse?
Cannot find it. Anyone know what to do in this situation?
Use Pimp My CPU instead.
I believe that Pimp my CPU is the only one working for our phone. You may want to check that out instead. If you search on XDA the Developer posted the app on here, however, I suggest you purchase the app from the market to support him.
Depends..
If your on stock rom and kernel then use setcpu and choose tegra2.
If your on custom rom with custom OC kernel use pimp my cpu.
Of course both require your phone to be rooted as well.
Sent from my LG-P999 using XDA Premium App
What exactly does SetCPU do that the stock froyo governor does not do?
Just curious.
This week I am trying a couple of things to see how big a difference Juice Defender and SetCPU really make.
^ Honestly haven't seen a difference with Juicedefender, unless I need to download the ultimate one. The basic one actually drains mybattery like no tomorrow.
Honestly, I use task killer but manual, I hardly kill my applications, I ignore most of them that need to be on or are on everytime I turn my phone.
But I have a girlfriend that likes to play 6 - 10 games at a time when she's on my phone, that needs to be killed when she's done or when I'm done with them.
I'm going to try SETCPU for tegra2 because I'm on stock right now. Looking to download a ROM though
What SETCPU does is it scales the processor frequency or speed in non techy terms.
For example:
If you enable screen off profile you can force the processor to run on slower speeds to conserve battery when the screen is off.
Recommended settings:
Max: 500
Min: lowest
Also with battery profile enabled if the phone goes below certain percentage it will drop the processor's speed to help conserve battery life as well but of course sacrificing performance though.
Recommended settings:
Max: lowest
Min: lowest
The main screen a.k.a general profile is the main profile the processor will be set to when the screen is on and the phone is being used.
Recommended settings:
Max: 1ghz
Min: 1ghz
What do you suggest for the other profiles such as charging or temp? Thanks for your post. It's the only one I've found in this thread with suggestions for the noobs.
mylomine said:
What do you suggest for the other profiles such as charging or temp? Thanks for your post. It's the only one I've found in this thread with suggestions for the noobs.
Click to expand...
Click to collapse
I don't have one for charging, but for temp, I have the max speed kick down to about half (512MHz or whatever the step value is) if the temp goes over 41C.
The most important one is the Screen Off profile, I have the max speed set to one step over minimum and it saves a lot of battery life.
sryan1983 said:
I believe that Pimp my CPU is the only one working for our phone. You may want to check that out instead. If you search on XDA the Developer posted the app on here, however, I suggest you purchase the app from the market to support him.
Click to expand...
Click to collapse
I beg to differ....cause I use setcpu.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Sent from my LG-P999 using XDA Premium App
Setcpu has worked for me on stock Froyo and stock Gingerbread, both rooted ofcourse.
Faux's kernel needs pimp, but morfic's Trinity kernel uses SetCPU and works great.
Sent from my Nexus 3 using XDA Ultimate App
LordButtersI said:
Faux's kernel needs pimp, but morfic's Trinity kernel uses SetCPU and works great.
Sent from my Nexus 3 using XDA Ultimate App
Click to expand...
Click to collapse
I've used SetCPU with Faux's kernel ever since the dev made SetCPU usable with Tegra devices.
In fact here is a post with Faux saying to use either http://forum.xda-developers.com/showpost.php?p=16056765&postcount=1695

[APP][2.2+] MonitorZ

MonitorZ
MonitorZ is a monitoring tool to gather infos about usage stats, resources and events at a desired time interval.
Data are provided into csv format @ sdcard/MonitorZ/ once a session is stopped.
No real time graphs to minimize overhead.
Mainly intended for developers
Stats monitors
These monitors provide a view on usage stats calculated on the selected time interval. cpu sleep time is always reported, stats are calculated on th "not sleeping time" of the each interval.
Scaling
This is the most innovative stats monitor. It provides a statistical measure of how cpu frequencies/states are used by the governor. The KPI is based on time spent on each state/frequency. This monitor gives 0 when 100% of the time is spent @ min frq and 100 when 100% of the time is spent @ max frq
Are provided two different alghoritms:
scaling A: it takes into account just the number of the available frequencies between min and max. It measures how governor goes through the differents available steps.
scaling B: it takes into account the number and the value of the available frequency between min and max. It measures how governor goes through the differents steps considering also their value and distribution between max and min.
cpu
standard cpu stats monitor
cpu + scaling
this monitor provides both scaling and cpu measure.
state
this monitor reports the time spent on each available cpu state frequency
Real time monitors
memory
standard memory monitor
battery
standard battery monitor
governor
this monitor reports the current governor
Event monitors
battery
standard battery monitor based on event (charge or % variation)
screen state
screen state monitor
airplane monitor
airplane status monitor
power connection
power connection status monitor
links
MonitorZ on market
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
Screen
Just thought I'd be nice and drop a working link here
https://market.android.com/details?id=com.chiabre.monitorz
version
1.1.1
scaling monitor
scaling + cpu monitor
label redesign
sdk to 2.2 for device compatibility
Click to expand...
Click to collapse
1.0.0
stop/play icon by andrea chiabrera
stats based monitors + deep sleep time %
cpu state monitor
cpu % + cpu state monitor aka governor intruder
memory monitor
Click to expand...
Click to collapse
0.9.8
governor monitor fix
sqlite backend
Click to expand...
Click to collapse
0.9.7
governor real time monitor
Click to expand...
Click to collapse
0.9.6
firmware version info fix
Click to expand...
Click to collapse
Hi all,
version 1.1.1 released.
AWESOME NEW SCALING MONITOR a fresh way to look into governors!
look @ first post 4 more infos!!!
Here some examples of what you can do with this monitor.
I did 3 test sessions on the 3 most used governor: conservative, performance e ondemand.
session detail:
- browser navigation for 2 minutes
- cpu+scaling monitor running @ 200ms.
session output:
- graph about cpu & scaling evolution
- avg governor efficiency based on cpu & scaling data
performance
avg efficiency 51%
conservative
avg efficiency 65%
ondemand
avg efficiency 75%
ANY COMMENT APPRECIATED

Kernel and vm messaround, potential lag fix

THOSE SETTING BELOW ARE SYSCTL, YEY I FOUND AMERICA
PROBABLY SOMEONE IS GOING TO LAUGH
​Hello,
The one and only thing I was bothering about is why some users can multitask great and I am always not able too.
The reason?
Well the apps are closing too frequently, going out of simple game makes the launcher load again, going back in game makes the game start all over again.
First I thought I need to limit myself and get used to it.
This is not a great multitasker as the iPhone is.
iOS is far from the best system but its multitasking capabilities amazed me.
So where is our problem? Why are we not able to multitask as we should?
Well in the past month I was messing around only with the RAM consumption.
I was aiming at more free RAM and persistent apps, here comes the V6 Supercharger which helps a lot.
But when I was enjoying it I got a call. What happend? Black screen and vibrating phone, that's all, I said to myself "No more playing, I have to answer my calls."
Where was the problem I believe I don't know still.
But with some dev help we might be able to control our phone a little bit more than we can right now with uboot and recovery. Thanks to all of the developers.
What was the problem? The Kernel! I installed ROM Toolbox and just got a look at the app when I saw the Kernel Tweaks.
The mess began, read which setting what does on vm and on kernel tweaks. Increased some, decreased others, tested it didint worked and I increased and decreased others and still the same when I saw this in vm
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
And
And this in kernel
Enabled them, decreased the swappiness (don't know if it has to be high, i increased it the first time to 100 but no change so I decreased it.)
And something happend, I called myself from another phone and there you go, no more black screen, I was able to answer while I was playing Shadowgun Deadzone at high graphics.
And the most amazing part was when I ended the call, the game was still running and I was connected.
In the process of testing I noticed a few freezes just in the launcher when opening apps for like 5 secs and will see what have I done, maybe I messed up something.
Anyways I was able to launch the task killer while the game was running in the background and I am so happy that it works fine for now.
If you read all this, thank you.
If you are developer and you think this is totally wrong I will remove this.
Hey I was wondering what was best settings I'm a bit of a noob when it comes to codes lol but I have a good understanding of it I think
Anyway in rom manager when it comes to performance I noticed io was set too noop and gov to interactive at this setting you get good performance but lags in games
So thinking from previous Roms for GB 21e most always io was sio and gov was hot plug
I'm not too sure what differences are between them in code but do see differences in performance
Any recommendations and maybe quick heads up would be highly appreciated
Running on Darkroom ICS int
Sent from my LG-SU760 using Tapatalk 2
MattJaO3D said:
Hey I was wondering what was best settings I'm a bit of a noob when it comes to codes lol but I have a good understanding of it I think
Anyway in rom manager when it comes to performance I noticed io was set too noop and gov to interactive at this setting you get good performance but lags in games
So thinking from previous Roms for GB 21e most always io was sio and gov was hot plug
I'm not too sure what differences are between them in code but do see differences in performance
Any recommendations and maybe quick heads up would be highly appreciated
Running on Darkroom ICS int
Sent from my LG-SU760 using Tapatalk 2
Click to expand...
Click to collapse
The setting you are using are perfect.
I was fiddling around with TWRP recovery and tweaking trough rom toolbox. I've set governer hotplug and I/O ondemand. SD boost on 4096kb. In TWRP I've created a swap file of 2GB and a EXT4 file of 2GB on my sdcard. I'm running ICS Darkroom international. In Sysctl Conf i set min free kb to 4096.
Phone is running very smooth and very very fast. I've a total new phone

Stock Kernel Optimisation

Hello M2 users, y'all probably wondering what did I mean by the thread title above so I'll get straight to the point nice and simple.
Using a stock kernel sucks, any CPU Tuner will go bonkers adjusting frequencies only to see it revert to stock again. So, while browsing for ways to optimise performance and battery in the Xperia, I came across this guide (now updated to this) on our Interactive governor. While it's on a different phone, the explanations are clear enough to tell anyone that yes, this can be adapted so long as you are rooted.
So read the whole post to give yourself an idea on the functions available for you. I use Kernel Adiutor for this as it seems most reliable.
But first, go into your build.prop files and edit this code:
Code:
ro.min_freq_0=300000
Set the frequency to 300000 (300Mhz) and restart or your phone will always be stuck on a higher frequency.
Also in the Kernel Samepage Merging, turn it off. There isn't much of an impact between battery and performance so this function is useless.
So now head into the CPU section, towards Tunables and add the following values according to the screenshot.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
[/img]
This gave me 50% battery usage over 26 hours, with 5 hours of Screen on Time used by chatting, gaming, and web surfing. Pretty good looking at all the times I've run the battery with less usage.
Anyhoo, I hope these settings are useful for y'all and if anyone has any modifications to these settings please do share it with your results. This may perhaps be the only guide where we control the governor on a deeper level rather than taking it as it is, so let's squeeze as much juice out of this phone as possible.
Update: These are the settings I've settled with, no negligible battery life and not much of an impact on performance. The voltage tables for the SD400 looks to be dynamic so you can't get an exact reading and anyone who looks at it will suggest you buy a new phone instead >.>
How do you add multiple values on one settings ?
I can't seem to be able to
FrWhyMe said:
How do you add multiple values on one settings ?
I can't seem to be able to
Click to expand...
Click to collapse
Right now you can only add multiples for above_hispeed_delay and target_load. Just tap the space bar each time you add in new values and the app will auto-arrange it for you. If it still doesn't work, what kernel are you using?
Can you repost the photo or write the values please?
basem_ym said:
Can you repost the photo or write the values please?
Click to expand...
Click to collapse
I'm not using this anymore, Sony's stock setting is fine unless you can somehow change the voltage.
AutumQueen92 said:
I'm not using this anymore, Sony's stock setting is fine unless you can somehow change the voltage.
Click to expand...
Click to collapse
I spent last week trying to study and understand those govenors.
Since yesterday, I was trying and tunning the stock interactive govenor
I didnt notice any major changes in performance but I was surprised by power consumption
Ex: I used to play Solitaire everyday more than 2 times a day, each time I spend like 1 hour and my battery drops like 10%
After tunning the interactive govenor my battery drops like 4-6%
Which means its much better than stock
I'm using EagleExperience rom which come with stock kernel
Anyway I just want to know your tunnables because the photo you posted is not opening
I want to compare them with my numbers
Thank you for this post and sorry for bad english
It's not my first lang.
basem_ym said:
I spent last week trying to study and understand those govenors.
Since yesterday, I was trying and tunning the stock interactive govenor
I didnt notice any major changes in performance but I was surprised by power consumption
Ex: I used to play Solitaire everyday more than 2 times a day, each time I spend like 1 hour and my battery drops like 10%
After tunning the interactive govenor my battery drops like 4-6%
Which means its much better than stock
I'm using EagleExperience rom which come with stock kernel
Anyway I just want to know your tunnables because the photo you posted is not opening
I want to compare them with my numbers
Thank you for this post and sorry for bad english
It's not my first lang.
Click to expand...
Click to collapse
These are my current settings, the 300Mhz and 787Mhz share the same voltage and any higher frequency has the same voltage too so it's more efficient to use Sony's settings. The only difference in here is I used Intelliplug to better control how they ramp up.
AutumQueen92 said:
These are my current settings, the 300Mhz and 787Mhz share the same voltage and any higher frequency has the same voltage too so it's more efficient to use Sony's settings. The only difference in here is I used Intelliplug to better control how they ramp up.
Click to expand...
Click to collapse
In target loads, you have frequencies that aren't supported on stock kernel (1.4GHz)
DavidMKD said:
In target loads, you have frequencies that aren't supported on stock kernel (1.4GHz)
Click to expand...
Click to collapse
That's because my kernel is overclocked.
above_hispeed_delay: 40000 998400:45000 1090400:50000 1190400:55000
boost: 0
boostpulse_duration: 0
go_hispeed_load: 99
hispeed_freq: 787200
io_is_busy_: 0
min_sample_time: 0
sampling_down_factor:1
sync_freq:300000
target loads: 80 600000:80 787200:90 998400:95 1090400:97 1190400:99
timer_rate: 60000
timer_slack: 480000
up_threshold_any_cpu_freq: 0
up_threshold_any_cpu_load: 0
Ksm off , no hotplug min:300 mhz max:1190 mhz
I am using this settings for now. I tried to lock 300-787 mhz for the my daily usage (whatsapp, youtube, cm browser, simple games like fun run) high frequences for heavy games, installing apps.
Is that config fine ? I tried semil's config its battery performance was great but i think it is slow config.
batuskas said:
above_hispeed_delay: 40000 998400:45000 1090400:50000 1190400:55000
boost: 0
boostpulse_duration: 0
go_hispeed_load: 99
hispeed_freq: 787200
io_is_busy_: 0
min_sample_time: 0
sampling_down_factor:1
sync_freq:300000
target loads: 80 600000:80 787200:90 998400:95 1090400:97 1190400:99
timer_rate: 60000
timer_slack: 480000
up_threshold_any_cpu_freq: 0
up_threshold_any_cpu_load: 0
Ksm off , no hotplug min:300 mhz max:1190 mhz
I am using this settings for now. I tried to lock 300-787 mhz for the my daily usage (whatsapp, youtube, cm browser, simple games like fun run) high frequences for heavy games, installing apps.
Is that config fine ? I tried semil's config its battery performance was great but i think it is slow config.
Click to expand...
Click to collapse
I try not to limit my frequencies since the point is to go high performance only when needed. I'm not sure about your config since I never turn off hotplugs.

[TESTING] Control Swappiness tendency

Hello guys, we have another new option in L Speed.
We are now able to control a swappiness tendency, how much aggressive we want to swap the memory.
{
"lightbox_close": "Close",
"lightbox_next": "Next",
"lightbox_previous": "Previous",
"lightbox_error": "The requested content cannot be loaded. Please try again later.",
"lightbox_start_slideshow": "Start slideshow",
"lightbox_stop_slideshow": "Stop slideshow",
"lightbox_full_screen": "Full screen",
"lightbox_thumbnails": "Thumbnails",
"lightbox_download": "Download",
"lightbox_share": "Share",
"lightbox_zoom": "Zoom",
"lightbox_new_window": "New window",
"lightbox_toggle_sidebar": "Toggle sidebar"
}
You can find this option under virtual memory section as you can see on this image. If there is any issue, please report.
Governor tuner doesn't work on this build.
Thanks for testing
Paget96 said:
Hello guys, we have another new option in L Speed.
We are now able to control a swappiness tendency, how much aggressive we want to swap the memory.
You can find this option under virtual memory section as you can see on this image. If there is any issue, please report.
Thanks for testing
Click to expand...
Click to collapse
Question: does this just change the swappiness percentage, that also can be changed in KA, or are other perimeters changed as well?
Thanks
martyfender said:
Question: does this just change the swappiness percentage, that also can be changed in KA, or are other perimeters changed as well?
Thanks
Click to expand...
Click to collapse
As you can see in L Speed logs, just swappiness.
Sent from my Nexus 5 using Tapatalk
Paget96 said:
As you can see in L Speed logs, just swappiness.
Sent from my Nexus 5 using Tapatalk
Click to expand...
Click to collapse
My 1GB device uses zram by default with a 400mb swap file and swapiness set to 100%. I have overcommit ratio set to zero. I have played with these settings in the past, but I'm not sure if I noticed any difference. Does lowering swapiness improve battery and cpu useage. Also what about lowering the amount of zram space? The only memory intensive app I have is GT Racing 2 Real Racing, but I rarely ever play it. Based on an xda guide, I have dirty_ratio set to 40 and dirty_background_ratio set to 10.
https://forum.xda-developers.com/android/general/guide-comprehensive-guide-to-kernels-t3121088
Thanks
martyfender said:
My 1GB device uses zram by default with a 400mb swap file and swapiness set to 100%. I have overcommit ratio set to zero. I have played with these settings in the past, but I'm not sure if I noticed any difference. Does lowering swapiness improve battery and cpu useage. Also what about lowering the amount of zram space? The only memory intensive app I have is GT Racing 2 Real Racing, but I rarely ever play it. Based on an xda guide, I have dirty_ratio set to 40 and dirty_background_ratio set to 10.
https://forum.xda-developers.com/android/general/guide-comprehensive-guide-to-kernels-t3121088
Thanks
Click to expand...
Click to collapse
It depends on whether you use swap/zram or no, lowering swappiness will reduce a "aggression" of how often kernel swap the memory. If it's lower, it's less, so less frequent CPU usage.
Paget96 said:
It depends on whether you use swap/zram or no, lowering swappiness will reduce a "aggression" of how often kernel swap the memory. If it's lower, it's less, so less frequent CPU usage.
Click to expand...
Click to collapse
I hadn't investigating vm changes much on this newer rom until yesterday. My device uses zram only, by default. Size 400mb, swapiness 100%. About 24 hrs ago I changed them to: size 250, swapiness, 40%. So far, my device appears to have less lag. Other settings I changed recently are dirty_ratio, 40, dirty_background_ratio,10, overcommit_ratio, 0, vfs_cache_ pressure remains at the system default of 100%
Do your cache reclaiming and cache allocation settings in lspeed, tune these parameters?
I/O tweak related: I had also enabled I/O boost and I/O extended que. I was experiencing some system lag, especially after wakeup, device unused overnight. Is it possible these two settings were the cause of the lag? I have them disabled now to test.
Thanks
martyfender said:
I hadn't investigating vm changes much on this newer rom until yesterday. My device uses zram only, by default. Size 400mb, swapiness 100%. About 24 hrs ago I changed them to: size 250, swapiness, 40%. So far, my device appears to have less lag. Other settings I changed recently are dirty_ratio, 40, dirty_background_ratio,10, overcommit_ratio, 0, vfs_cache_ pressure remains at the system default of 100%
Do your cache reclaiming and cache allocation settings in lspeed, tune these parameters?
I/O tweak related: I had also enabled I/O boost and I/O extended que. I was experiencing some system lag, especially after wakeup, device unused overnight. Is it possible these two settings were the cause of the lag? I have them disabled now to test.
Thanks
Click to expand...
Click to collapse
Yes, L Speed tune those parameters. And yeah, those can result a lag, test it and report to me.
Since we don't have anything to complain about this option, I assume we don't have to test it anymore. You are pleasured with those parameters, so I'll close this thread.
Thread re-opened, waiting for some reports about this option and potential improvements.
Thanks for support
We are also fine with this option, nothing much to test here.
If you have any problems with this option, please write to this thread.
Thanks for support

Categories

Resources